Vasodilators such as dipyridamole, lidoflazine and phentolamine produce a steal phenomenon (decreased collateral blood flow) in the heart in which collateral circulation has developed secondary to a coronary occlusion. Halothane and morphine have vasoactive properties. Hence both halothane and morphine may be expected to affect the collateral circulation. Though halothane and morphine-nitrous oxide anesthesia are frequently used in surgical patients with ischemic heart disease, their effects on collateral blood flow have not been determined. We propose to measure this in a canine model of experimentally induced myocardial ischemia. Ameroid constrictors will be placed around the left circumflex coronary artery of dogs to produce chronic total occlusion in 2-3 weeks. Inflatable cuff occluders will be placed around left anterior descending coronary artery to produce stenosis at the time of the experiment. Four weeks after instrumentation effects of halothane and morphine-nitrous oxide anesthesia on regional myocardial blood flow to the nonischemic area and collateral blood flow to the ischemic area will be determined using the radioactive microsphere technique. Measurements during anesthesia will be made without and with partial stenosis of the left anterior descending coronary artery in the presence of an already occluded left circumflex coronary artery and compared to awake control measurements. The information obtained should help provide a clearer rationale for anesthetic management of patients with ischemic heart disease.